Numerous technologies for forming flat-panel displays are known in the art. One such technology is the electro-luminescent display, which is formed by coating a thin layer of electro-luminescent material between a pair of electrodes. Displays employing this technology produce light as a function of the current between the two electrodes when the electro-luminescent materials are electrically stimulated. Electro-luminescent displays are primarily classified as active-matrix or passive-matrix displays. Active-matrix displays employ a relatively complex, active circuit at each pixel in the display to control the flow of current through the electro-luminescent material layer(s). The formation of this active circuit at each pixel can be expensive and often the performance of these circuits is somewhat limited. For example, when controlling current to a light-emitting element, circuits provided in low temperature polysilicon often exhibit spatial nonuniformities while circuits provided in amorphous silicon often exhibit severe threshold shifts over time.
Passive-matrix EL displays are much simpler in their construction. The display generally includes an array of row electrodes and an array of column electrodes. EL materials are deposited between these electrodes, such that when a positive electrical potential is created between the two electrodes, the EL material between these two electrodes emit light. Therefore, each light-emitting element in the display is formed by the intersection of a row and a column electrode. As this type of display does not require the costly formation of active circuits at each pixel site, they are much less expensive to construct. In these devices, the column electrode is typically formed of ITO or some other material that is transparent but typically higher in resistivity than the row electrode, to allow light to be visible to the user.
Numerous passive matrix EL display systems have been described in the literature. For example Okuda et al. in U.S. Pat. No. 5,844,368, entitled “Driving system for driving luminous elements” describes a system for driving a passive matrix EL display. In this method, and in most traditional passive matrix EL drive methods; it is assumed that a power is provided to one row electrode at a time and flows through the EL material to each of the column lines. This method of driving the display by providing power to only one line of light-emitting elements leads to two significant problems.
The first of these two problems, occur because each display will ideally have hundreds of lines of light-emitting elements, which implies that each light-emitting element will only emit light for a very short period of time. Therefore each light-emitting element will be required to emit light with a very high luminance to achieve a reasonable time-averaged luminance value. Since light intensity from these devices is proportional to current, relatively high currents must be provided to each light-emitting element. This can significantly shorten the lifetime of the individual light-emitting elements and increase cross-talk between pixels in the display as described by Soh, et al, in a paper entitled “Dependence of OLED Display Degradation on Driving Conditions” and published in the proceedings of the SID Mid Europe Chapter in 2006. Further this drive method requires drive electronics to support high currents, which usually translate to larger, more expensive silicon drive chips; and leads to high resistive voltage and power losses across the electrodes, especially the row electrodes which provide current to potentially hundreds of light-emitting elements simultaneously.
The second of these two problems occur because each light-emitting element must be turned on and off during each cycle to avoid current leakage, and therefore light emission, through light-emitting elements that are supposedly not activated. This problem is particularly troubling in EL displays employing organic materials since the EL layers are very thin and are highly resistive. In such displays, each light-emitting element has a significant capacitance that must be overcome before light emission can occur. Overcoming this capacitance can require significant power that does not generate light and is therefore wasted. This issue has been discussed by Yang et al. in a paper entitled, “PMOLED Driver Design with Pre-charge Power Saving Algorithm” as published in the 2006 SID Digest. As this paper states, this power increases significantly as the number of lines in the display is increased. Specifically, this paper points out that for a PM OLED having 64 lines, nearly 80% of the power is spent driving the OLED (i.e., for light production), while 20% of the power is spent overcoming this capacitance as the lines are turned on and off. As the resolution increases, this ratio changes dramatically, such that when there are 176 lines, only 57% of the power is spent in the production of light while 43% of the power is spent overcoming this capacitance. Therefore, the display becomes significantly less energy efficient, as more lines are present on the display to be cycled from off to on.
Each of these problems can significantly limit the use of passive matrix EL displays. However, in combination, these two problems limit the application space for such displays significantly. Today, the application of passive matrix EL displays are limited to displays that generally have less than 128 lines and are typically less than 1.5 inches in diagonal.
One category of approaches for addressing at least the first of these two problems is to provide multi-line addressing of passive matrix EL displays. Such methods have the potential to reduce the peak current through any EL light-emitting element, which can extend the lifetime of the material and significantly reduce the drive voltage. Further, since multiple rows can be engaged simultaneously, the power losses due to the resistivity of the electrodes can be reduced significantly.
In US Patent Publication No. 2004/0125046, entitled “Image Display Apparatus”, by Yamazaki et al., one such multi-line addressing method is provided. While disclosed primarily for use in surface-conduction type electron emitting devices, this approach was also been discussed for EL displays. In this approach, any input image signal that has fewer vertical addressable pixels than the vertical addressability of the display is displayed by receiving the input video signal, providing a horizontal edge emphasis process (i.e., edge sharpening) across the column direction of the display, selecting two or more rows of the display, and modulating the voltage to the columns of the display in response to the processed input image signal. This approach requires relatively straightforward image processing to prepare the image signal and is able to employ drivers that are very similar to existing passive matrix drivers. While this method may reduce the drive current and voltage as compared to a display employing one line at a time drive techniques as known in the prior art, simply providing the same signal on two neighboring lines, results in an image with a substantial loss in sharpness in the vertical direction and the edge emphasis process can provide only a limited level of enhancement. Therefore, while it is possible to use this method to provide a relatively good display when simultaneously selecting two rows of the display at a time and under certain circumstances it may be useful to select three rows at a time, the number of rows that can be employed simultaneously without introducing significant levels of image blur is quite limited.
Sylvan in EP 1 739 650, entitled “Procédé de pilotage d'un dispositif d'affichage d'images à matrice passive par selection multilignes” has proposed an enhancement to this method in which multiple rows are selected during one refresh of the display but a single row is selected during subsequent display refresh cycles. This approach overcomes at least a portion of the sharpness issues but requires that the display actually be cycled more often, further increasing the number of charge and discharge cycles and therefore increasing the power to capacitance. In a paper entitled “Multiline Addressing by Network Flow” by Eisenbrand et al., a similar approach has also been discussed. This approach allows some cycles to be completed using even more rows simultaneously but employs a hierarchical approach that once again requires the use of an increased number of charge and discharge cycles.
A different approach has more recently been discussed by Smith et al. in PCT filings WO 2006/035246 entitled “Multi-line addressing methods and apparatus”, WO 2006/035248 entitled “Multi-line addressing methods and apparatus” and WO 2006/067520 entitled “Digital Signal Processing Methods and Apparatus”. These disclosures provide a method for decomposing an input image into subframes, using mathematical methods such as singular value decomposition and then displaying these subframes by controlling multiple rows and columns in an emissive display simultaneously. An interesting difference between this approach and the prior approaches is that the prior approaches provided only a single scan signal value to the selected row columns and typically provided a digital time multiplexed signal to the columns. The approach provided by Smith requires that multiple drive levels be provided on both the column and row electrodes. In fact, the method as described requires full analog control over the signals provided on the row and column electrodes and possibly requires that the current to each of these electrodes be controlled. While this adds complexity to the drivers, it also allows more control that can be used to engage more rows simultaneously with fewer artifacts. Unfortunately, the methods described in each of the disclosures by Smith, suffer from a number of shortcomings. Most importantly, the decomposition methods described are complex and difficult to realize in real time, especially when processing full frames of video information.